The latest from John Hussman: Markets are wildly overvalued by any reasonable measure, and you’re not nearly getting compensated enough to jump in now.
Basically, stocks only look like they’re an OK buy because margins are extremely high, and analysts have unrealistic projections.
With any kind of mean reversion, or using any other kind of valuation technique (such as smoothed earnings), stocks are ridiculously overvalued.
We’ve regularly observed that corporate profit margins (and economy-wide, profits as a share of GDP) have a strong tendency to “mean revert” over time – specifically, elevated profit margins are associated with unusually weak earnings growth over the following 5-year period, and depressed profit margins are associated with unusually strong earnings growth over that horizon (see last week’s comment, A False Sense of Security ). Notably, the ratio of corporate profits to GDP is presently nearly 70% above its historical norm. Of course, the most common valuation methods used by Wall Street analysts (whether they use the “Fed model” or “forward operating earnings times arbitrary P/E multiple”) rely almost exclusively on estimates of year ahead earnings. Embedded in these toy models is the quiet assumption that current profit margins will be sustained indefinitely.
By contrast, a wide range of measures that use “normalized” fundamentals of one form or another are extraordinarily stretched. Andrew Smithers recently took note of the elevated levels of cyclically adjusted P/E ratios and price to replacement cost (“q”) and observed “As of 8th March, 2012, with the S&P 500 at 1365.9 , the overvaluation by the relevant measures was 48% for non-financials and 66% for quoted shares. Although the overvaluation of the stock market is well short of the extremes reached at the year ends of 1929 and 1999, it has reached the other previous peaks of 1906, 1936 and 1968.”
At 1400 on the S&P 500, the market’s overvaluation has now reached 70% on these measures, which have a far stronger correlation with subsequent market returns than the Fed Model or other unadjusted methods using forward operating earnings. This is particularly true over horizons of 4 years or longer. As a side note, since the reliance on forward operating earnings is now an established Wall Street practice, Valuing the S&P 500 Using Forward Operating Earnings details how to improve the reliability of market valuations based on these figures.
We presently estimate a nominal total return on the S&P 500 averaging 4.1% annually over the coming decade. This modestly exceeds the yield available on a 10-year Treasury, but by a small margin that – outside the late 1990’s bubble period – has previously been seen only during the two-year period approaching the 1929 peak, between 1968-1972 (which was finally cleared by the 73-74 market plunge), and briefly in 1987, before the crash of that year.
While it’s true that interest rates are depressed, apparently setting a low “bar” for equities, an additional question one should ask is whether interest rates themselves are “fair” in the sense of being adequate compensation for long-horizon risks. For example, back in 1982, stocks had a reasonable 10-year prospective risk-premium versus bonds, but both were priced to achieve extraordinarily strong returns. Presently, stocks have a weak 10-year prospective risk-premium versus bonds, but both are priced to achieve unsatisfactory returns. In 1982, investors had an incentive to lock in either, and were served well regardless of their choice. At present, investors have no reasonable incentive at all to “lock in” the prospective returns implied by current prices of stocks or long-term bonds (though we suspect that 10-year Treasuries may benefit over a short horizon due to continued economic risks and still-unresolved debt concerns in Europe, which has already entered an economic downturn).